Résumé
Background: available data are insufficient to determine the ideal duration of antibiotic therapy. The concept of post-aggression acquired immunodepression in intensive care is a well-established concept, and its assessment with a decreased monocytic HLA-DR (mHLA-DR) was described in several studies. Hence, we aimed to evaluate the prognostic value of this marker prior to discontinuing antibiotic therapy on the risks of recurrence, superinfection, or mortality.Design, setting, and participants: a prospective monocentric study in an intensive care unit (ICU) of a tertiary hospital in France from September 2022 to August 2023. The study included critically ill adults admitted for all causes and requiring antibiotic therapy for at least 7 days. Upon ICU admission, epidemiological data, infection type, antibiotic therapy, comorbidities, and Sequential Organ Failure Assessment (SOFA) score were collected.Standardized quantitative evaluation of mHLA-DR was performed at baseline (at day 3) and weekly thereafter. Patients were followed up for the development of secondary infections 40 days after antibiotic therapy discontinuation (ATD).Main outcome and measures: the primary endpoint was a composite outcome including the occurrence of recurrence, superinfection, or death at day 40 after ATD. Secondary outcomes were to compare the mHLA-DR kinetics to other biomarkers and SOFA score, and to assess the causes of antibiotic therapy failure.Results: among the 257 patients enrolled in the study, 118 patients experienced at least one ATD, and 48% of them were surgical patients. Patients with consistently low mHLA-DR levels (below 8,000 Ab/c twice) before discontinuation of antibiotic treatment had a significantly higher risk of experiencing the primary outcome (p< 0.0037).Multivariate Cox analysis revealed that mHLA-DR below 8,000 Ab/c twice before ATD, presence of candida positive sample, presence of multidrug-resistant bacteria (MDR), SOFA score >5 without decrease of the SOFA score before ATD, significantly impact the occurrence of death, superinfections, and relapse. Based on these findings, a score was developed yielding an AUC (area under the curve) of 0.78 (95% CI: 0.67–0.87), which can be used to identify the low-risk population for antibiotic therapy failure.Conclusion: monocytic HLA-DR expression appears to be a good marker of acquired immune deficiency following aggression, and its kinetics before ATD could help determining the optimal duration of antibiotic therapy in the era of personalized medicine. Nonetheless, this assessment must be integrated with the clinical history of each patient, and further studies are needed to support its relevance in clinical practice.